Borderline Personality Disorder (BPD) is a psychiatric disorder defined, in part, by recurrent impulsive and suicidal behaviors. With a suicide rate of up to 10%, BPD is among the most lethal of psychiatric disorders, comparable to affective and schizophrenic disorders. This study is the continuation of a first prospective, multidimensional study of the clinical, psychosocial and biological risk factors for suicidal behavior in subjects ascertained for BPD. The overarching goal of this study is to identify clinically meaningful predictors of suicidal behavior, and subgroups of BPD patients at highest risk of suicide completion. Results of this study will guide clinical risk assessment and identify targets for treatment intervention. Upon entering the study, subjects are assessed for pre-defined risk factors, including demographic and socioeconomic variables, personality trait vulnerabilities, psychiatric comorbidities, substance use disorders, childhood histories of abuse, social adjustment and treatment histories. Annual follow-up assessments provide prospective, longitudinal data on suicidal behaviors, psychiatric and substance use disorders, life events, and interim treatment efforts. We will identify risk factors which predict time-to-attempt in short (1 yr, 2 yr.), intermediate (2-5 yr.) and long-term time frames (5 - 10 yrs.) using appropriate proportional hazards regression models. We will also characterize clinically recognizable BPD subgroups at highest risk of suicidal behavior using Latent Class Analysis. Trajectory analysis will identify subgroups with increasing medical lethality of attempts over time. Identifying risk factors and subtypes associated with recurrent attempts and increasing lethality of attempts over time addresses an urgent clinical need to identify patients on a trajectory toward suicide completion. We will also test theoretical models for the etiology of suicidal behavior in BPD. Using mediation analysis, we will examine the relationships between adverse childhood experiences (e.g. sexual, physical abuse), traits of temperament (e.g. impulsivity, aggression), clinical risk factors (e.g. MDD, ASPD), and adult suicidal behavior. The psychobiology of suicidal behavior will be addressed through PET neuroimaging studies of the serotonin-2A receptor, which has been implicated in post-mortem studies of suicide victims, and an MRI study of structural brain changes associated with childhood abuse in BPD. Neuroimaging data will be tested with other clinical variables for significance in predicting suicidal behavior in an effort to identify biologic "markers" in individuals at highest risk